Dieter Davis

CX Operating Model

A structured customer experience function inside a large matrix organization. Two teams serve every line of business, geography, brand and channel: one runs the platforms and reacts, the other mines them and goes looking for money.

Two sheets, US Letter landscape. Print this page directly (landscape, no margins, background graphics on) or download the PDF. Companion reference: Stages of CX Operational Excellence.

Dieter Davis · Customer Experience Analytics · Sheet 1 of 2

A structured CX team inside a large matrix organization

Two CX teams serve every line of business, geography, brand and channel: one runs the platforms and reacts, the other mines them and goes looking for money.
THE BUSINESS MATRIX each dot is a product team with a product manager Line of Business A Line of Business B Line of Business C … Geographies Brands Digital & Retail channels Business & Technology business / tech owners REACTIVE LANE PROACTIVE LANE Requests & self-service new surveys, replay searches, tagging, issues escalated by Customer Service Tools embedded at launch answers, root cause, monitoring, one trusted CX dataset for everyone Sized $ opportunities, A/B tests, fixes plus a win report each team carries to its own leadership PM partnership test slots, deploys, shared credit CX SWAT PROACTIVE Special Weapons & Tactics. Mines the platforms for recurring complaints, sizes them as conversion lift or cost savings, and proves the value landed. 2 formed when 1 is at 100% integration asks, new signals to capture quant + qual data, platform access CX Analytics PLATFORM · REACTIVE · OPERATIONS Runs the CX tool stack as a shared service: the matrix self-serves or files a request; every launch ships with the tools in place. 1 formed first Session replayquant e.g. HCL Discover what customers did, at scale Listening platformsqual e.g. Qualtrics, Verint what customers said, and why Central repository one consolidated CX dataset replaces redundant licenses Deploy & integration tools in every product, correctly consistent, reliable releases
Vertical position shows build order: the platform layer is established before the proactive layer is staffed. Arrow color shows which lane a flow belongs to.

Build the foundation first. CX Analytics absorbs every request and makes the data trustworthy. Proactive work staffed before that point gets pulled into reactive tickets.

Two lanes, one dataset. The reactive lane answers what the business asks. The proactive lane brings the business what it did not know to ask for. Both read from the same platforms.

Wins belong to the business. SWAT hands each product team a documented win to report up its own chain. Credit flows through the matrix, which is why teams keep partnering.

Dieter Davis · Customer Experience Analytics · Sheet 2 of 2

The two teams: what they run, how they work, who they serve

Numbers match the build order on sheet 1. Vendor names are examples; the roles are what matter.
1 · CX AnalyticsPlatform, reactive & operations

Mandate

Run the CX tool stack as a shared service so any part of the matrix can see and respond to customer problems from one trusted dataset.

What they run

  • Session replay (quant): HCL Discover or equivalent. Behavior at scale, struggle, drop-off.
  • Listening platforms (qual): Qualtrics, Verint. Surveys, intercepts, open-text feedback.
  • Central repository: one consolidated CX dataset, one license footprint.
  • Integration & deploy standards: tags, SDKs and survey placements in every LOB product, in the right places, released the same way every time.

How they work

  • Self-service first. Dashboards, replay search and survey builders are open to the matrix inside governance rules.
  • Request queue with an SLA for anything beyond self-service: new surveys, integrations, deeper analysis.
  • Reactive triage. Issues raised by Customer Service or a leader are reproduced in replay, explained with feedback, and routed to the owning team with root cause.
  • Embedded at launch. Every new product or release includes CX tooling in the launch plan, both instrumentation and go-live monitoring.
  • Consolidation. Redundant licenses are retired; deploys converge on one pattern.

Partners

Product teams in every LOB, brand, geography and channel. Technology for deployment, data and security. Customer Service as the loudest early-warning signal. Procurement for license consolidation.

Measures of success

Share of products instrumented. Request SLA attainment. Time from reported issue to root cause. License spend retired. Adoption of the central dataset by the business, which is the test of trust.

2 · CX SWATProactive. Special Weapons & Tactics

Mandate

Use the platforms to find, size and prove money the business is leaving on the table, before anyone asks. Formed only once CX Analytics is handling 100% of inbound requests.

The operating loop

  1. Mine. Scan listening data for recurring complaints; confirm the behavior in session replay.
  2. Quantify. Size the affected population and convert it to dollars: conversion lift, or operational cost savings such as avoided contacts.
  3. Partner. Bring the sized opportunity directly to the product manager who owns the experience.
  4. Test or fix. A/B test the proposed solution, or deploy an outright fix when the answer is clear.
  5. Monitor. Measure after release and confirm the value landed. No confirmed value, no win.
  6. Report. Hand the team a win deliverable to carry to its own leadership; log it on the wins dashboard; roll it into the year-end summary.

Partners

Product managers, the primary relationship. CX Analytics for data, platform access and new integrations. Experimentation and optimization teams for test design. Finance to validate the dollar figures. Customer Service for the complaint signal.

Measures of success

Validated dollars per year, split by conversion lift and cost savings. Wins per quarter. Cycle time from finding to live test. Number of product teams actively partnered. Year-end summary delivered to the executive team.

The sequencing rule

Analytics before SWAT. SWAT findings are only credible when the data is trusted and deploys are consistent. A proactive team staffed before that point spends its time on reactive requests and never reaches the money.

Self-serve

Any team in the matrix uses dashboards, replay search and survey tools directly, inside governance set by CX Analytics.

Request

Anything beyond self-service goes through the queue with an SLA: new surveys, integrations, deeper analysis, escalations from Customer Service.

Launch

Every new product, release or market entry carries CX tooling and go-live monitoring in its launch plan. No launch is instrumented after the fact.